非線性系統之智慧型適應控制系統設計

Intelligent Adaptive Control System Design for Nonlinear Systems

冷金緒(Chin-Hsu Leng)

指導教授 ： 林志民

元智大學/電機通訊學院/電機工程學系/博士(2008年)

摘要

在實際系統控制過程中，廣泛存在非線性性質、複雜的干擾及各種不確定性，而傳統控制理論方法往往較難以勝任；因此，本論文針對一些非線性系統控制問題，研究智慧型控制理論方法，主要基於「適應控制技術」與「滑動模式(Sliding Mode)控制」架構，設計不同模糊邏輯與類神經網路之適應控制器，提昇系統控制性能。
在適應模糊滑動控制(AFSMC)系統控制器設計方面，提出積分型與比例積分型兩種模糊滑動模式參數調控方法；另外，為能解決對未知系統的模型誤差問題，將適應模糊滑動控制(AFSMC)系統擷取模式誤差資訊回饋調整適應控制法則，設計模糊識別適應滑動控制(FIASMC)系統架構，可改善控制誤差。
在類神經網路系統控制器設計方面，利用類神經網路容錯、平行計算及線上參數學習特性，提出兩種適應性類神經網路強健控制系統，分別為 控制補償之輻射基本函數型(RBF)類神經網路及 控制補償之回饋型類神經網路(RNNC)系統，均能藉由類神經網路之參數收斂調整及強健穩定補償控制，獲致良好追蹤性能；最後，結合模糊推論與類神經網路兩者之優點，研析混合式模糊類神經滑動模式控制方法，設計適應回饋型模糊類神經網路(ARFNN)控制器，模仿理想的滑動模型，並完善強健補償控制，具誤差修正及類神經網路自我線上訓練學習能力，其性能較優於比例積分型模糊滑動控制系統。
本論文所提出的模糊邏輯與類神經網路控制器，均利用李亞普諾夫穩定理論或最陡坡降法調整法則，增加網路收斂速度與系統之穩定性；同時，亦將其控制理論方法，分別應用於Van der Pol振盪器、壓電陶瓷伺服馬達、機翼震盪、聲波馬達及直流交換式電源轉換器等系統，以驗證這些方法之可行性。

關鍵字

智慧型；適應控制：模糊；類神經網路

並列摘要

Unknown uncertainties, perturbations, disturbances and nonlinearities often exist in practical systems and many traditional control system design methodologies can not be applied in these systems. This dissertation is mainly devoted to and focused on the design of intelligent adaptive controllers based on the sliding-mode and adaptive control technologies for uncertain nonlinear systems.
For the adaptive fuzzy sliding-mode control (AFSMC) system design, two types of AFSMC system controllers are developed to speed up the convergence of tracking errors by adjusting the controller parameters with the use of a set of fuzzy rules; one is an integral AFSMC (I-AFSMC), and the other is a proportional-integral AFSMC (PI-AFSMC). For both of the controllers, some parameters are tuned either by an integral learning algorithm or by a proportional-integral learning algorithm. Furthermore, in order to solve the AFSMC system modeling error problem, a fuzzy-identification-based adaptive sliding-mode control (FIASMC) scheme is proposed and is used to bring the modeling-error information into the adaptive laws so as to achieve better control performance.
Neural networks (NNs) have characteristics of fault tolerance, parallelism, and on-line learning capability. For the neural network controller design, two robust adaptive neural network control systems are developed. One is a radial basis function (RBF) neural network to approximate the unknown system dynamics with the control technique; in this approach the tracking error can be attenuated to a specified level. The other is a robust neural network control (RNNC) system to achieve trajectory tracking performance based on the neural network approach with the control technique. Finally, a fuzzy neural network which incorporates the advantages of fuzzy inference and neuron-learning in designing an adaptive recurrent fuzzy neural network (ARFNN) controller is developed. An ARFNN is used to mimic an ideal sliding mode controller and a compensation controller is designed to recover the residual of the approximation errors. The trained ARFNN controller that is applied to an on-line learning algorithm can thus achieve better performance than the PI controller.
In these intelligent control system design, the on-line parameter tuning methodology using both of the Lyapunov stability theorem and the gradient descent method is developed to increase the system learning capability and to provide better system stability. The developed control system design methods are then applied to some control system applications (such as Van der Pol oscillator, linear piezoelectric ceramic motor, chaotic Duffing system, wing rock system, Chau’s chaotic circuit system, Linear ultrasonic motor (LUSM), and DC-DC converters) for demonstrating the effectiveness of the proposed design methods.

並列關鍵字

Intelligent ；Adaptive Control ：fuzzy：Neural networks

參考文獻

[3] C. C. Lee, “Fuzzy logic in control systems: fuzzy logic controller-part I/II,” IEEE Trans. Systems, Man, and Cybernetics, vol. 20, no. 2, pp. 404-435, 1990.

[4] H. O. Wang, K. Tanaka, and M. F. Griffin, “An approach to fuzzy control of nonlinear systems: stability and design issues,” IEEE Trans. Fuzzy Systems, vol. 4, no. 1, pp. 14-23, 1996.

[5] R. Palm, “Robust control by fuzzy sliding mode,” Automatica, vol. 30, no. 9, pp. 1429-1437, 1994.

[6] S. W. Kim and J. J. Lee, “Design of a fuzzy controller with fuzzy sliding surface,” Fuzzy Sets and Systems, vol. 71, no. 3, pp.359-369, 1995.

[7] X. Yu, Z. Man, and B. Wu, “Design of fuzzy sliding-mode control systems,” Fuzzy Sets and Systems, vol. 95, no. 3, pp. 295-306, 1998.